On Sat, 24 Apr 2004 22:26:49 +0000, elmeras2000
<
jer@...> wrote:
>You may be told right away that they correspond to the Sanskrit
>vr.kí:h.-type with which -id- was identified by Chantraine. My wife
>Birgit Olsen has made a spectacle of herself by assuming that IE *-
>iH2-o- gave Greek *-ido-, so that Gk. gen. -ídos equals Ved. -ías.
>The intermediate stage supposedly had something like -iDo- with a
>dental spirant (much like Welsh -ydd from -iyo-). I'm afraid this
>reflects my bad influence on her. The paper was published in an
>Erlangen congress report (Indoarisch, Iranisch und die
>Indogermanistik, Wiesbaden 2000). Nobody liked it, but there were no
>arguments against it.
I'd have to take a look at the Greek and Vedic part of the
argument. I just checked up on the Armenian side of the
argument, summarized on p. 852 of "The Noun in Biblical
Armenian":
... the occurrences of a "suffixal" *-d- > -t- somehow
appear connected with "eRu/Ru:"-roots where "*d" takes the
position of the laryngeal which would be root final in the
zero grade (*Rhu > *Ruh, cf. Rasmussen 1989): alawt
"obscure, concealed" < *pl.ud- < /pluh2-/, arawt 'pasture,
pasturage' < *sr.ud- < /sruh2-/, cnawt (-ic`) 'jaw; volute'
< *g^enud- <- /g^enuh1-/ and karawt 'needing' < *gWr.ud- <
/gWruh2-/...
One thing I find curious about these forms is the syllabic
resonant (*pl.ud- *sr.ud-, *gWr.ud-). If there had been
simple metathesis of laryngeal and /u/ in *plh2u-, *srh2u-,
*gWrh2u-, I would expect *pluh2, *sruh2-, *gWruh2-, not
*pl.uh2- *sr.uh2-, *gWr.uh2-. If there's an explanation,
I'd like to hear it, but for now I prefer to assume that
*pl.u- *sr.u-, *gWr.u- normally reflect *pl.h2u- *sr.h2u-,
*gWr.h2u-. So what is the *d?
I can understand nobody liked it: a development from a
voiceless (post-)velar fricative to a voiced dental stop is
rather hard to swallow. But assuming the Greek forms in
*-id- are indeed connected to the Skt. vr.ki:s type, what
other explanation can there be?
One thing about Greek -id- is that has a short /i/. That
may perhaps indicate that *h2 was absorbed by a following
consonant. The logical choice would be *-t, which appears
as a final and apparently meaningless element in Vedic
root-nouns ending in -u, -i and -r (jí-t, s'rú-t, kr'-t).
This element apparently does not ablaut (at least in Vedic).
Could it have been added to *-ih2? We would have:
nom. *-íh2-t-z > *-ids
acc. *-íh2-t-m > *-idm.
gen. *-ih2-t-és > *-idés, etc.
Whether *d instead of expected, perhaps, *th in the oblique
forms was extended from the strong forms, or that we can
have *-ih2t- > *-id- directly, I wouldn't know.
=======================
Miguel Carrasquer Vidal
mcv@...